Marketing Analytics

Optimize your marketing investments with reliable data

• Marketing Mix Modeling (MMM): Measure the real impact of each channel on sales and optimize budget allocation
• Multi-touch Attribution: Understand the complete customer journey and value each touchpoint
• Campaign Analysis: Evaluate campaign performance with actionable KPIs (ROAS, CPA, LTV)
• Customer Segmentation: Identify your most profitable segments to personalize actions

Technologies: R (Robyn, Prophet), Python, BigQuery, Looker Studio

See MMM project →

Data Engineering

Build a modern and scalable data infrastructure

• ETL/ELT Pipelines: Automate data extraction, transformation, and loading
• Data Warehousing: Design optimized cloud architectures (BigQuery, Snowflake)
• dbt Transformation: Implement versioned, tested, and documented transformations
• Streaming Analytics: Process real-time data with Kafka and Spark

Technologies: dbt, BigQuery, Airflow, Kafka, Spark, Delta Lake, Docker

See Pipeline project →

Experimentation & A/B Testing

Make decisions based on evidence, not intuition

• Experiment Design: Create statistically robust tests with appropriate power calculations
• Bayesian Analysis: Get interpretable results with credible intervals
• Causal Inference: Go beyond correlation to measure real impact of your actions
• Heterogeneous Effects (CATE): Identify segments that respond best to your interventions

Technologies: PyMC, CausalML, DoWhy, statsmodels, ArviZ

See Causal Inference project →

Applied Machine Learning

Deploy models that create business value

• Predictive Models: Anticipate churn, demand, sales with interpretable models
• Dynamic Pricing: Optimize prices in real-time based on context and demand
• Recommendation: Personalize customer experience with recommendation systems
• MLOps: Deploy and monitor production models (Vetiver, FastAPI, Docker)

Technologies: scikit-learn, XGBoost, PyTorch, Vetiver, FastAPI, MLflow

See Dynamic Pricing project →

1. Discovery

Together we define your business objectives, identify available data and technical constraints. I propose an approach adapted to your context.

2. Development

I develop the solution iteratively, with regular checkpoints to validate direction. You have complete visibility on progress.

3. Delivery

I deliver a documented, tested, and ready-to-use solution. I support you in getting started and remain available for support.